Robust planning with (L)RTDP

Olivier Buffet; Douglas Aberdeen

Robust planning with (L)RTDP

Olivier Buffet, Douglas Aberdeen

Research output: Contribution to journal › Conference article › peer-review

19 Citations (Scopus)

Abstract

Stochastic Shortest Path problems (SSPs), a subclass of Markov Decision Problems (MDPs), can be efficiently dealt with using Real-Time Dynamic Programming (RTDP). Yet, MDP models are often uncertain (obtained through statistics or guessing). The usual approach is robust planning: searching for the best policy under the worst model. This paper shows how RTDP can be made robust in the common case where transition probabilities are known to lie in a given interval.

Original language	English
Pages (from-to)	1214-1219
Number of pages	6
Journal	IJCAI International Joint Conference on Artificial Intelligence
Publication status	Published - 2005
Event	19th International Joint Conference on Artificial Intelligence, IJCAI 2005 - Edinburgh, United Kingdom Duration: 30 Jul 2005 → 5 Aug 2005

Cite this

@article{464dccc7266c450890e68b32d777fc4c,

title = "Robust planning with (L)RTDP",

abstract = "Stochastic Shortest Path problems (SSPs), a subclass of Markov Decision Problems (MDPs), can be efficiently dealt with using Real-Time Dynamic Programming (RTDP). Yet, MDP models are often uncertain (obtained through statistics or guessing). The usual approach is robust planning: searching for the best policy under the worst model. This paper shows how RTDP can be made robust in the common case where transition probabilities are known to lie in a given interval.",

author = "Olivier Buffet and Douglas Aberdeen",

year = "2005",

language = "English",

pages = "1214--1219",

journal = "IJCAI International Joint Conference on Artificial Intelligence",

issn = "1045-0823",

publisher = "International Joint Conferences on Artificial Intelligence",

note = "19th International Joint Conference on Artificial Intelligence, IJCAI 2005 ; Conference date: 30-07-2005 Through 05-08-2005",

}

TY - JOUR

T1 - Robust planning with (L)RTDP

AU - Buffet, Olivier

AU - Aberdeen, Douglas

PY - 2005

Y1 - 2005

N2 - Stochastic Shortest Path problems (SSPs), a subclass of Markov Decision Problems (MDPs), can be efficiently dealt with using Real-Time Dynamic Programming (RTDP). Yet, MDP models are often uncertain (obtained through statistics or guessing). The usual approach is robust planning: searching for the best policy under the worst model. This paper shows how RTDP can be made robust in the common case where transition probabilities are known to lie in a given interval.

AB - Stochastic Shortest Path problems (SSPs), a subclass of Markov Decision Problems (MDPs), can be efficiently dealt with using Real-Time Dynamic Programming (RTDP). Yet, MDP models are often uncertain (obtained through statistics or guessing). The usual approach is robust planning: searching for the best policy under the worst model. This paper shows how RTDP can be made robust in the common case where transition probabilities are known to lie in a given interval.

UR - http://www.scopus.com/inward/record.url?scp=78650598019&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:78650598019

SN - 1045-0823

SP - 1214

EP - 1219

JO - IJCAI International Joint Conference on Artificial Intelligence

JF - IJCAI International Joint Conference on Artificial Intelligence

T2 - 19th International Joint Conference on Artificial Intelligence, IJCAI 2005

Y2 - 30 July 2005 through 5 August 2005

ER -

Robust planning with (L)RTDP

Abstract

Other files and links

Fingerprint

Cite this